By adapting the rating mechanism, readers can preliminatively judge the quality of the paper. Only the subject closely related reviewers can rate. Deep Science will recommend and select most closely related reviewers by using keywords match up.

The rating will be judged by four levels: Quality, repeatability, innovation and impact.

Reviewers will provide a rating from one star (poor) to five stars (excellent)

Is the publication of relevance for the academic community and does it provide important insights? Is the language correct and easy to understand for an academic in the field? Are the figures well displayed and captions properly described? Is the article systematically and logically organized?

Is the hypothesis clearly formulated? Is the argumentation stringent? Are the data sound, well-controlled and statistically significant? Is the interpretation balanced and supported by the data? Are appropriate and state-of-the-art methods used?

Does the work represent a novel approach or new findings in comparison with other publications in the field?

Does the work have potential huge impact to the related research area?

There is an overall average rating for a preprint paper. An average rating for each different level will also be recorded. There is no limit on the number of ratings.